Experimental Setup

The experimental setup used in this work consists of an open loop subsonic wind tunnel with a test section of 100×100×750 mm. The test section is made of glass in order to resist against the high corrosion of biodiesel. The environment air is blown into the wind tunnel by a 1.5HP blower fan made by Aeroflo (Mississauga, ON, Canada). Then the flow passes through a fine screen and a nozzle before reaching the test section. The velocity inside the test section can be varied between 20 and 58 m/s by means of using a damper located at the inlet of the fan. The maximum air velocity in the test section is characterized by Particle Image Velocimetry (PIV) measurements of a very fine spray (less than 10μm) in order to find the exact velocity vector directions, turbulence intensity, and to calibrate the measurements of a Pitot tube. The PIV measurements (see Figure

‎2.16, Appendix at page 50) show that in the test section, the velocity of air is parallel and constant at each axial location except the narrow boundary layers on the wall, which are

28

less than 10 mm thick on each side. Therefore, the air axial velocity can be considered uniform in the liquid jet’s upstream. In addition the PIV results show a turbulence intensity of 9% exists at the wind tunnel test section just upstream of the injector location.

After the calibration of the Pitot tube for the maximum velocity with the PIV measurements, the air velocity and pressure inside the test section are measured using the Pitot tube. A schematic of the liquid injector is shown in Figure ‎2.1. As can be seen in Figure ‎2.1, the injector is a plain circular tube with a diameter of 0.5 mm. A tapered transition from the 2 mm to 0.5 mm is applied to avoid cavitations. The length to diameter is also considered 100 in order to have fully developed flow in all cases, even the lowest velocity, which can cause the laminar flow inside the injector; however, the high-pressure losses by using this long injector are compensated for by the fuel pump.

The injector is mounted vertically on the top of the test section at an axial position of 200 mm from its inlet plane. The center of the 0.5 mm diameter lies on the symmetry plane of the test section.

Figure ‎2.1 Schematic of the wind tunnel test section and injector

Shadowgraph with a high-speed camera (Photron SA1.1, USA) has been used for capturing images of the spray side view (x-y plane in Figure ‎2.1). In order to have images with good contrast, even for the smallest droplets on the windward edge of the spray, the

200mm

29

trajectory was captured at 250 frames per second and the shutter speed was set at 1/593000 sec. Having this configuration, 500 images, with a resolution of 1024×1024, were captured for each test. In addition, 500 images were captured from the background without any liquid spray. In order to eliminate the background noise from the spray images, and digitizing the images to find the spray windward trajectory, a code was written using the image processing toolbox of MATLAB. The minimum value of each pixel, among 500 background images, creates the background image. The next step in the image processing code is to eliminate the background image from the spray image. The resultant images without background noise are superimposed to form the average spray trajectory. A sample raw image and the superimposed image of this process are illustrated in Figure ‎2.2. Using a 25% light cut-off threshold [47], the windward and leeward spray boundaries are defined as the loci of the points with the lowest light intensity. Finally, a number of points (i.e., 50-100) are sampled on the locus of windward trajectory and a curve is fitted to the sample points. It should be mentioned that the uncertainty of the obtained trajectory is proportional to the size of image, the image resolution and the number of pixels per jet diameter. Having about 11 pixels per diameter and the size of image (100×100 diameter), lead to 0.1% uncertainty in the trajectories.

30

Figure ‎2.2 a) Snapshot of diesel spray, We = 80, q=82, b) Filtered and averaged 500 snapshots of diesel spray, We = 80, q=82, c) Snapshot of B100 spray, We = 76, q=102, d)

Filtered and averaged 500 snapshots of B100 spray, We = 76, q=102. Circle markers indicate the windward trajectory points of the spray and the lines are exponential fitted

curves. Reference points are the windward trajectory points 50 mm downstream

Two-dimensional Phase Doppler Particle Analyzer (PDPA) by TSI Inc. (MN, USA) is used as a vehicle for other experiments. Two components of the droplet velocity, as well as the droplet diameter, are captured in a 143° forward scattering of the PDPA probes (TSI Inc.). The PDPA is set up on a traverse with a 0.1 mm spatial increment (ISEL Germany). The focusing point, considering the refractions of the glass walls, is

(a) (b)

(c) (d)

Ref

Ref Ref

Ref

10 mm

10 mm

31

found and set on the reference point at a 50 mm section downstream of the nozzle exit. It should be mentioned that, for each test, the reference point is selected to lie on the windward trajectory point (see Figure ‎2.2). In addition, three other points are captured under this point (see Figure ‎2.3) in order to make sure that this point is lying on the windward trajectory. In order to capture the whole spray characterization, 25 points in a 5×5 matrix are set up to perform the velocimetry and capture the droplet size distribution.

Furthermore, to find out that the 5×5 matrix lies in the spray and does not reach the leeward location of the spray, five more points are set up on the top of the matrix. The PDPA tests are conducted on this matrix for diesel, biodiesel, and different blends 50 mm downstream of the injection point. It should be mentioned that a droplet sphericity measurement is performed according to the work of Araneo et al. [48], for several traverse planes downstream the orifice. The test resulted in having higher than 90%

spherical droplets among the total captured droplets at 50 mm downstream, while the same test depicted for example ~65% spherical droplets/ligaments at 25 mm plane downstream that may translate to inaccurate measurement or incomplete secondary breakup. Consequently, 50 mm plane downstream the orifice have been selected for PDPA measurements.

32

Figure ‎2.3 Test Section Schematic, the planes on which PDPA is done and the matrix of capturing points

Diesel, biodiesel (B100) and several blends (i.e., B50, B20 and B5) are tested for several momentum flux ratios and Weber numbers. The blends are named based on the percentage of the mass ratio of biodiesel to diesel. For example, B20 is made by 20% by mass of biodiesel added to 80% by mass of diesel. The biodiesel in the present work is an ASTM 6751-based vegetable fatty acid methyl ester which is refined and produced by Rothsay Biodiesel, Canada, from animal fat and recycled cooking oil. In addition, this diesel is the known D-2 type diesel provided by Ultramar, Canada. The physical properties of biodiesel (B100) and biodiesel blends B20 and B5 are obtained mainly from the material fact sheet of the same day, same batch production and also from the literature [49], [50] which has a similar combination. In fact, a comparison between the repeated physical properties from the literature and the factsheet shows a negligible ±5%

33

difference. The properties are shown in Table ‎2.1. Additionally, the properties of B50 are interpolated between the properties of B20 and B100.

Table ‎2.1 Physical properties of diesel, biodiesel, and blends

Properties Unit Diesel B100 B50 B20 B5

The test matrix for each liquid is composed of three different air velocities. For each air velocity, three momentum flux ratios are tested. Consequently, 45 different test conditions are performed to include a wide range of momentum flux ratios (q) and Weber numbers (Table ‎2.2). The test conditions for diesel and biodiesel blends are adjusted to provide the same amount of heat of combustion for all test cases. For example, the momentum flux ratio of 100 for diesel has an equal heat of combustion as the momentum flux ratio of 135 for B100, if we assume the same combustion efficiency. This results in having higher mass flow rates for the biodiesel blends compared to those of the diesel case, but this unbiased test condition seems more practical in the industrial applications, where there are different choices of fuel for the same combustion chamber.

Table ‎2.2 Experimental test conditions

34